Since the aging population is growing at an unprecedented pace, the number of Veterans diagnosed with dementia is estimated to increase dramatically. Identification of individuals before the onset of significant clinical symptoms is essential for facilitating intervention when therapies may be most effective. One approach to identifying potential biomarkers of prodromal dementia involves assessing brain structure in individuals at increased risk for developing Alzheimer's disease (AD), such as those with mild cognitive impairment (MCI). With a recent explosion in MCI research, it has become clear that it is a heterogeneous disorder and distinct clinical subtypes (e.g., amnestic, nonamnestic) have been proposed. Heterogeneity in the clinical presentation of MCI may relate in part to the heterogeneity of underlying neuropathological mechanism(s). Although debated, evidence suggests that individuals with amnestic MCI demonstrate greater medial temporal lobe damage typical of early AD (e.g., Singh et al., 2006). In contrast, those with nonamnestic MCI show greater frontal lobe involvement (Nobili et al., 2008) and greater cerebrovascular disease (as evidenced by white matter abnormalities) on MRI (Delano-Wood et al., 2009). Mixed MCI characterized by impairment in multiple domains of cognitive ability may represent individuals with mixed pathology with both neurodegenerative and cerebrovascular features (Libon et al., 2010). Taken together, previous studies highlight a complex relationship between MCI, white matter, AD pathology, and cerebrovascular functioning. The proposed study will combine imaging markers of white matter abnormalities and vascular risk markers with AD-related biomarkers (e.g., cerebrospinal fluid [CSF] measures of amyloid and tau, hippocampal volume) as well as take into account MCI clinical subtypes in order to more completely characterize the contributions of AD and cerebrovascular risk to the dementia prodrome. In doing so, we can compare the prevailing model of biomarkers of MCI and AD, which emphasizes neurodegeneration (e.g., Jack et al., 2013) with a model relating the combined effects of vascular and AD pathologies to the probability of developing dementia (Chui et al., 2012). The current study aims to (1) better characterize the structural brain changes and vascular risk profiles underlying distinct MCI subtypes, (2) critically examine novel neuroimaging measures that might distinguish MCI subtypes and contribute to the nature and severity of cognitive impairment, and (3) determine which neuroimaging markers of white and gray matter alterations and additional risk factors for dementia (i.e., vascular risk factors, APOE 4 status) are most useful in predicting cognitive decline and progression to dementia. In the proposed study, 92 older adults (23 cognitively normal, 23 amnestic MCI, 23 nonamnestic MCI, and 23 mixed/multiple domain MCI) will undergo comprehensive neuropsychological assessment; laboratory testing to assess vascular risk factors (e.g., fasting glucose); CSF measurement of amyloid and tau; and neuroimaging exams including high resolution structural, perfusion, and white matter imaging. White matter imaging protocols include T2-weighted fluid attenuated inversion recovery (FLAIR) as well as two cutting edge techniques: diffusional kurtosis imaging (DKI) and a novel, myelin-selective technique (multi-component driven equilibrium single pulse observation of T1 and T2 [mcDESPOT]). It is hypothesized that greater vascular risk burden and white matter changes in frontal regions will be associated with nonamnestic MCI and mixed/multiple domain MCI. In contrast, white matter changes in medial temporal lobe regions will be associated with amnestic MCI and mixed/multiple domain MCI. A better understanding of the underlying neuropathology associated with distinct MCI clinical subtypes and the implementation of multimodal neuroimaging markers that better detect heterogeneous pathologies may have important implications for diagnosis, prognosis, treatment selection, and monitoring of disease-modifying effects of therapy as well as selection criteria for clinical trials.